Search results

[ncclx][c10d] check if process group nccl is using nccl

fb-exported

meta-exported

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 163160

#163160 In pytorch/pytorch;

· riklas opened on Sep 17, 2025

[xpu] Support high stream for ProcessGroupXCCL

ciflow/xpuRun XPU CI tasksRun XPU CI tasks

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

open source

release notes: distributed (c10d)release notes categoryrelease notes category

release notes: xpurelease notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 163049

#163049 In pytorch/pytorch;

· Chao1Han opened on Sep 16, 2025

Revert "Make distributed modules importable even when backend not bui…

ci-no-tdDo not run TD on this PRDo not run TD on this PR

ciflow/h100-symm-mem

ciflow/inductor

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

oncall: jitAdd this issue/PR to JIT oncall triage queueAdd this issue/PR to JIT oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

release notes: inductor (aoti)

Status: Open (in progress).

pytorch/pytorchnumber 163024

#163024 In pytorch/pytorch;

· Camyll opened on Sep 15, 2025

[RFC]: No Distributed Log Spew

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 162999

#162999 In pytorch/pytorch;

· msaroufim opened on Sep 15, 2025

[POC] Print backtraces of where unwaited work was allocated

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 162979

#162979 In pytorch/pytorch;

· ezyang opened on Sep 15, 2025

Add option to FakeProcessGroup to raise error if comms are invoked.

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 162841

#162841 In pytorch/pytorch;

· ezyang opened on Sep 12, 2025

[c10d] destroy_process_group() works without default group created

ciflow/trunkTrigger trunk jobs on your pull requestTrigger trunk jobs on your pull request

keep-goingDon't stop on first failure, keep running tests until the endDon't stop on first failure, keep running tests until the end

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 162770

#162770 In pytorch/pytorch;

· kwen2501 opened on Sep 11, 2025

[WIP] LocalTensor

ciflow/inductor

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 162753

#162753 In pytorch/pytorch;

· ezyang opened on Sep 11, 2025

[SymmMem] Add <code>num_active_allocations</code> and dealloc checks

ciflow/h100-symm-mem

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 162681

#162681 In pytorch/pytorch;

· kwen2501 opened on Sep 11, 2025

[SymmMem] Fix memory allocation hold-up

ciflow/h100-symm-mem

ciflow/trunkTrigger trunk jobs on your pull requestTrigger trunk jobs on your pull request

merging

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

topic: bug fixestopic categorytopic category

Status: Open (in progress).

pytorch/pytorchnumber 162680

#162680 In pytorch/pytorch;

· kwen2501 opened on Sep 11, 2025

[DDP] Rebuild gradient buckets when computation graph changes

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

open source

release notes: distributed (c10d)release notes categoryrelease notes category

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open (in progress).

pytorch/pytorchnumber 162667

#162667 In pytorch/pytorch;

· whwjiang opened on Sep 10, 2025

Back DTensor with Symmetric Memory ops

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

release notes: distributed (c10d)release notes categoryrelease notes category

Status: Open (in progress).

pytorch/pytorchnumber 162662

#162662 In pytorch/pytorch;

· kwen2501 opened on Sep 10, 2025